Back to Search
Start Over
The RheaG Weather Generator Algorithm: Evaluation in Four Contrasting Climates from the Iberian Peninsula.
- Source :
- Journal of Applied Meteorology & Climatology; Jan2019, Vol. 58 Issue 1, p55-69, 15p, 3 Charts, 9 Graphs
- Publication Year :
- 2019
-
Abstract
- This paper describes the assumptions, equations, and procedures of the RheaG weather generator algorithm (WGA). RheaG was conceived for the generation of robust daily meteorological time series, whether in static or transient climate conditions. Here we analyze its performance in four Iberian locations—Bilbao, Barcelona, Madrid, and Sevilla—with differentiated climate characteristics. To validate the RheaG WGA, we compared observed and generated meteorological time series' statistical properties of precipitation, maximum temperature, and minimum temperature for all four locations. We also compared observed and simulated rain events spell length probabilities in all four locations. Finally, RheaG includes two weather generation procedures: one in which monthly mean values for meteorological variables are unconstrained and one in which they are constrained according to a predefined baseline climate variability. Here, we compare the two weather generation procedures included in RheaG using the observed data from Barcelona. Our results present a high agreement in the statistical properties and the rain spell length probabilities between observed and generated meteorological time series. Our results show that RheaG accurately reproduces seasonal patterns of the observed meteorological time series for all four locations, and it is even able to differentiate two climatic seasons in Bilbao that are also present in the observed data. We find a trade-off between generation procedures in which the unconstrained procedure better reproduces the variability of monthly and yearly precipitation than the constrained one, but the constrained procedure is able to keep the same climatic signal across meteorological time series. Thus, the first procedure is more accurate, but the latter is able to maintain spatial autocorrelation among generated meteorological time series. [ABSTRACT FROM AUTHOR]
- Subjects :
- CLIMATE change
METEOROLOGICAL precipitation
RAINFALL
CLIMATOLOGY
ALGORITHMS
Subjects
Details
- Language :
- English
- ISSN :
- 15588424
- Volume :
- 58
- Issue :
- 1
- Database :
- Complementary Index
- Journal :
- Journal of Applied Meteorology & Climatology
- Publication Type :
- Academic Journal
- Accession number :
- 134550823
- Full Text :
- https://doi.org/10.1175/JAMC-D-18-0170.1